Blame French Theory? In this Economy?

Notes on fully automated culture war

Last week, the (wonderfully named) French tech bro Brivael Le Pogam set X alight with an apology on behalf of his nation for having loosed “French Theory” on the world. The post was in French, but the phrase was in English, as was the case with François Cusset’s 2005 book on the subject; for Francophones, use of the English phrase points to the Americanized version of Gallic thought. Here is how he began:

From there, he went on to say, these amoral, nihilistic ideas—exusable, perhaps, as language games played over espressos at Les Deux Magots—were taken up by Americans, who vulgarized and moralized them, birthing that monstrous changeling, le woke-isme.

It’s a familiar story, so much so that many have suggested the tweet was LLM-generated, plausible both given its derivative content, formulaic syntax, and the fact that M. Le Pogam works in AI (more on that below). Regardless, a few things were striking about the post and the response to it. The first is how big an audience there is for yet another reiteration of this shopworn tale: It was retweeted by Elon, Bill Ackman, and other X luminaries, and at current count has received 55 million views. Clearly, many are eager to hear, yet again, the sordid tale of how a bunch of degenerate Frenchies are to blame for present cultural woes.

More noteworthy than the ever-freshness of hackneyed narratives of foreign intellectual subversion was Le Pogam’s conclusion, a bracing encomium to his own industry:

In other words: French Theory deconstructed Truth, Beauty, and Goodness, and now the tech industry is going to rebuild them. “Au travail!” was the post’s rousing finale.

Now, let’s have a look at just what M. Le Pogam was getting back to work on after he signed off. His company, Argil, was founded in 2023 and has received funding from the vaunted startup accelerator Ycombinator. What bold project of cultural renewal is Argil undertaking, you ask? Surely something of monumental civilizational importance…:

Yes, you have that right. The “things” Le Pogam’s company is “making, instead of deconstructing” are … tools for “generating videos with your AI clone.” This is how he is restoring Truth, Beauty, and Goodness to our culture, thereby doing penance for his country’s malign intellectual influence: by adding to the burgeoning global overcapacity of slop generation machinery. What could save the West from civilizational collapse faster than you being able to instantly generate videos of yourself dressed as an Irish mobster?:

Snark aside, something interesting is going on here. For the cultural conservatives who denounced Foucault and co. decades ago in similar terms, the source of Truth, Beauty, and the Good to which we had to return was the Western canon, Christianity, the family, and so on. Ditch that fashionable nonsense, read the classics, go to church, vote for Reagan. But apparently, for today’s anti-woke firebrands, a tool for making a personalized AI clone is just as powerful an engine of civilizational renewal as the sublime glories of Notre Dame and Chartres.

To be clear, the remarkable thing about this claim isn’t that the proposed tools are patently inadequate defenses against the West’s slide into decadence, but that they seem—pretty obviously!—to contribute to the same degeneracy that is being denounced. Indeed, it doesn’t require more than a few seconds of reflection to recognize that the technologies Le Pogam promotes as the counterpoint to French Theory produce many of the things he attributes to the latter. Consider the following scandalous “French Theory” theses he enumerates:

• “texts have no stable meaning”

• “every signifier slips away”

• “the author is dead”

Ooh la la! Le Pogam attributes these shocking aphorisms to Derrida, although the “death of the author” is Barthes and the “slippage of the signifier” is Lacan (which suggests his own rigor about proper attributions of authorship may suffer from the same defects as most other internet users, a point not irrelevant to the original “death of the author” thesis…). Anyway, it is remarkable to declare yourself scandalized by the assertion that “the author is dead” when you work in the AI industry. Just this week, the literary world is abuzz over the fact that a seemingly AI-generated text, whose author’s public photo seems to also be AI-generated, was awarded the Commonwealth Prize and published in Granta. Mass use of LLMs has operationalized the death of the author on a scale unimaginable when French Theorists issued their provocations on the subject. What Barthes declared in 1967—“the text is a tissue of citations, resulting from the thousand sources of culture”—is a simple statement of fact about the products of generative AI. For more on these points, I refer readers to Leif Weatherby’s Language Machines.

Let’s turn to another of Le Pogam’s examples of harmful French Theory ideas:

• “every identity is constructed and thus negotiable”

Again, it’s quite astonishing for a guy whose company markets what it calls “AI fictions,” which it promises will allow users to “become the main character and put yourself in ANY situation from a single picture, in minutes” to also enlist himself in the defense of the stability and fixity of identity. Likewise, when Le Pogam laments “an entire generation [that] knows how to suspect and no longer knows how to admire [and] sees power everywhere and beauty nowhere,” one has to marvel at the cognitive dissonance required for someone marketing tools that enable people to create and circulate “AI clones” of themselves so as to compete for status and power to then lament that the young folks today only “know how to suspect” and not to “admire.”

To figure out what’s going on here, it is worth revisiting a canonical work of French Theory, Jean-François Lyotard’s The Postmodern Condition: A Report on Knowledge. Lyotard famously defined postmodernism as “incredulity towards metanarratives”—a definition that encompasses the skepticism about Truth, Beauty, and Goodness decried by Le Pogam and like-minded culture warriors over the past five decades. What is the source of this incredulity? Did a handful of French philosophers produce it all on their own, as Le Pogam and so many culture warriors before him have suggested? Lyotard’s answer may seem surprising: “this incredulity is … a product of progress in the sciences.” Specifically, as he argues in his first chapter, postmodernity is a specific condition affecting “knowledge in computerized societies.”

How could this be? Don’t science and technology seek to establish Truth, not undermine it? To address this objection, Lyotard distinguishes between the pragmatic, experimental pursuit of facts within technoscientific research and the broader cultural dimension of Truth, which depends on narrative. “Science,” Lyotard writes, “has always been in conflict with narratives.” This is because within the internal workings of experimental research, all that is needed is to produce provisional results, not an overarching narrative about their value and meaning. Only when scientists need to justify themselves to the public does science “legitimate itself [by] making an explicit appeal to some grand narrative, such as the dialectics of Spirit, the hermeneutics of meaning, the emancipation of the rational or working subject, or the creation of wealth.” The grands récits whose demise Lyotard declared were ways of situating localized facts within a broader account of Truth that endowed them with comprehensive significance.

In the late twentieth century, however, computerization drove a wedge between science and narrative through its “exteriorization of knowledge with respect to the ‘knower.’” The idea is that rather than being something stored inside human brains and shared between human subjects, scientific knowledge has turned into something produced and circulated through impersonal technological systems. Lyotard writes: “The old principle that the acquisition of knowledge is indissociable from the training (Bildung) of minds, or even of individuals, is becoming obsolete and will become ever more so.” Bildung was the unfolding of the individual mind in relation to knowledge: an individual story that mirrored the larger narrative of scientific progress. But as computerization permitted new paradigms to take hold (Lyotard cites genetics as a key example), knowledge production became automated and distributed across vast informational networks.

Compounding this situation, computerization massively increases the volume of specialized knowledge, ensuring that a single mind is no longer able to hold more than a minuscule fraction of the sum of knowledge. Knowledge is thus further displaced from minds and into systems, and “each of us lives at the intersection of many of these.” Truth has dispersed into truths. For specialized researchers, what matters isn’t forging a broad consensus as to the truth-value of a particular finding, but generating results that are useful within a particular sphere of activity.

What Lyotard offers here is a concrete, material, and institutional account of what a French Theorist might call the decentering of the subject. As computer systems advance, the human knower becomes incidental to the transactions of knowledge production. This is the process the AI industry now seeks to bring to completion, replacing human experts with, in the words of Anthropic’s Dario Amodei, “a country of geniuses in a data center.” Artificial General Intelligence is the endpoint of the “computerization” process identified by Lyotard.

The material and institutional transformations described by Lyotard are the background conditions that made postmodern ideas plausible, accessible, and appealing in the late twentieth century. In a world in which knowledge production unfolded across a vast machinery far beyond the scope of any individual mind, ideas about the disappearance of the author as guarantor of meaning, the pure instrumentality of truth, the topology of the rhizome, and so on offered a means of “cognitive mapping,” to use Frederic Jameson’s phrase in Postmodernism.

The account I am offering here relies on a loosely Marxian model of base and superstructure, where transformations of the means of production—in this case, technoscientific knowledge production—precede and bring about shifts in consciousness. In other words, we became postmodernists—we took up French Theory—because we already lived in the postmodern condition. Our ideas about truth shifted because the processes by which truths were being produced and circulated already had done so previously. Likewise, well before LLMs, the technological mediation of literary and artistic production was deconstructing the mythology of the author or individual human creator; in the realm of art, Marcel Duchamp and Andy Warhol were well ahead of this game, as were the early surrealist pioneers of automatic writing.

This account allows us to understand right-wing culture war polemic as a fetishistic enterprise. Because disruptions at the level of the base were felt as symptoms at the level of the superstructure, right-wing culture warriors could assign blame on superstructural actors and elements, thereby obscuring underlying material causes. As I wrote about previously, this is roughly A.J.A. Woods’s account of the “cultural Marxism conspiracy theory”: “Without a structural understanding of capitalist crises,” Woods writes, cultural conservatives “needed to locate some sort of external cause for social disintegration,” and “cultural radicalism was the perfect scapegoat.”

If Le Pogam’s culture war salvo is anything to go by, though, the accelerated phase of postmodernity we have entered is also destabilizing the base-superstructure distinction relied upon above. The technologies Le Pogam is unleashing on the world are cultural ones, as Weatherby argues. This means that, as we have seen, they directly and immediately produce the exact effects he wants us to blame on “French Theorists.” Today, Le Pogam, Musk, and their ilk are themselves the real “cultural radicals” of the current moment, overseeing the destabilization of truth, the self, beauty, and so on far vaster scale and a more rapid pace than any experimental artist or post-structuralist intellectual could ever have dreamed of. With this closing of the loop, the two phases coincide: the transformations of the “base” enacted by generative AI already operate instantly at the level of superstructure.

It is as if Le Pogam were standing over the murdered corpses of Beauty, Truth, and the Good, smoking gun hot in his hand, telling the gendarmes: “I know who did this! It was French Theory!” But the absurdity of the scene risks distracting from what is really going on. I think Le Pogam, Musk, et al. really believe that they are on the side of the corpses they are helping to bury, of civilization, the West, and other supposedly solid things. But they are able to believe this because the postmodern condition has long since deprived these terms of any underlying reality. They are signifiers that have slipped away and now float across a cultural battlefield that has become a collection of automated gestures and habits, a clone of itself, an “AI fiction.”

This week in Compact:

• William Jones on Pope Leo’s AI encyclical

• Ralph Leonard on American empire’s durability

• Jared Hayden on the GUARD Act

• Daniel McCarthy on Thomas Massie’s loss

• Paul Gottfried on what he learned from Herbert Marcuse

• Stephen Adubato on Gen Z’s superficial religiosity

• Ernest Jesuyemi on literary safetyism and AI fiction